Many decisions require the retrieval of relevant information from memory. Thus, it is important to understand how decision-making and memory processes interact with each other. Gluth and colleagues (Gluth, Sommer, Rieskamp, & Buechel, 2015) showed that people prefer remembered over forgotten options even when the forgotten option was more likely to be better. Further analyses suggested that people seem to compare recalled items to a subjective reference value, that replaced the value of the forgotten option and that was below the average value of all options. This implies that people’s choices were biased by their memory. One possible mechanistic explanation for the existence of this bias is that people might believe that they remember options of high value better than options of low value. In this case, it would be rational to reject forgotten options more frequently. In the current project, we aim to test three hypotheses: First, we seek to replicate the memory bias reported in Gluth et al. (2015). Second, we ask whether people believe that they are more likely to remember options of higher value as compared to options of lower value. Third, we test whether the memory bias and the “believe bias” are linked to each other: Do people who believe more strongly that they remember better options better also exhibit a stronger memory bias?